Application of Loewner framework for data-driven modeling and diagnosis of polymer electrolyte membrane fuel cells

Abstract

Environmental and resources problems have been the driving force behind the development of fuel cell technologies. The polymer electrolyte membrane fuel cell (PEMFC) is one of the highly promising fuel cells in terms of delivering energy requirements for a vast number of applications. Nevertheless, its commercialization has been restricted because of its limited durability and reliability. In order to enhance its performance, effective modelling, and diagnostic strategies are essential. Several technologies are employed to investigate the various degradation mechanisms occurring in the PEMFC. Among them, electrochemical impedance spectroscopy (EIS) is the most widely employed method. Nevertheless, it is not able to distinguish processes having a similar time constant, and thus alternative frequency response analysis (FRA) techniques have recently been developed involving non-electrical inputs and/or outputs, for instance, the concentration-alternating frequency response analysis (cFRA). However, these methodologies have required a longer period of experiments and their results are difficult to interpret, which requires complex models to understand them. In order to tackle such challenges, the data-driven based approach, known as the Loewner Framework, is adopted in this thesis. Applying the Loewner framework, a new methodology is developed as a complementary analysis technique for interpreting EIS and cFRA data of the PEMFC. This novel method allows the identification of the different features of the individual physicochemical phenomena in a very clear manner and facilitates the decoupling of processes with comparable time constants. In addition, the cFRA experiment data are analysed by using the Loewner framework in order to shorten the duration of the experiments, and the results show the feasibility of a significant amount of time reduction of the cFRA experiments.